AI agents will exploit weak cyber habits regardless of system age
Tech experts warn that replacing legacy government systems won't stop AI attacks without fixing poor cyber habits first.
The federal government has ordered agencies to audit old computer systems. New plans to reduce their use must be drawn up by March 31, following last week's data hack by an OpenAI agent.
Tech leaders praised the directive but warned it misses the core problem. A tech firm executive compared it to fixing an old bank vault when the real problem is someone left it unlocked.
The core issue is that government departments have ignored basic cyber security for years, experts say. AI agents are ruthless and fast at finding even obscure weaknesses.
Older systems may even be more secure because they have had decades to fix known problems. What matters now is making sure cyber controls are solid everywhere, not rushing to buy new hardware.
- March 31 next year
- Audit deadline
- OpenAI agent
- Attacker
- Department of Home Affairs
- Issuing department
- Protective Security Policy Framework
- Policy framework
Why it mattersA cyberattack on government systems can expose millions of Australians' personal data and disrupt services like welfare and tax systems.
Open this story in InSnip →
